13C-Isotope Dansylation Labeling and Fast Liquid

Mar 23, 2009 - Kevin Guo and Liang Li*. Department of Chemistry, University of Alberta, Edmonton, Alberta T6G 2G2 Canada. We report a new .... Page 2 ...
2 downloads 0 Views 308KB Size
Download PDF
Anal. Chem. 2009, 81, 3919–3932

Differential 12C-/13C-Isotope Dansylation Labeling and Fast Liquid Chromatography/Mass Spectrometry for Absolute and Relative Quantification of the Metabolome Kevin Guo and Liang Li* Department of Chemistry, University of Alberta, Edmonton, Alberta T6G 2G2 Canada We report a new quantitative metabolome profiling technique based on differential 12C-/13C-isotope dansylation labeling of metabolites, fast liquid chromatography (LC) separation and electrospray ionization Fouriertransform ion cyclotron resonance mass spectrometry (ESI-FTICR MS) detection. An isotope reagent, 13Cdansyl chloride, can be readily synthesized. This reagent, along with 12C-dansyl chloride, provides a simple and robust means of labeling metabolites containing primary amine, secondary amine, or phenolic hydroxyl group(s). It is shown that dansylation labeling offers 1-3 orders of magnitude ESI signal enhancement over the underivatized counterparts. Dansylation alters the chromatographic behaviors of polar and ionic metabolites normally not retainable on a reversed phase (RP) column to an extent that they can be retained and separated by RPLC with high efficiency. There is no isotopic effect on RPLC separation of the differential isotope labeled metabolites, and 12 C-/13C-labeled isoforms of metabolites are coeluted and detected by MS for precise and accurate quantification and confident metabolite identification. It is demonstrated that, in the analysis of 20 amino acids, a linear response of over 2 orders of magnitude is achieved for relative metabolite quantification with an average relative standard deviation (RSD) of about 5.3% from replicate experiments. A dansylation standard compound library consisting of 121 known amines and phenols has been constructed and is proven to be useful for absolute metabolite quantification and MS-based metabolite identification in biological samples. As an example, the absolute concentrations of 93 metabolites, ranging from 30 nM to 2510 µM, can be determined from a pooled sample of human urines collected in 5 consecutive days labeled with 12C-dansylation and spiked with the 121 13Cdansylated standards. Relative concentration variations of these metabolites in individual urine samples can also be monitored by mixing the 13C-dansylated pooled urine sample with the 12C-dansylated individual sample. With a 12 min fast LC separation combined with FTICR MS, 672 metabolites were detected in a human urine * Corresponding author. E-mail: [email protected]. 10.1021/ac900166a CCC: $40.75  2009 American Chemical Society Published on Web 03/23/2009

sample with each metabolite peak having a signal-tonoise ratio of greater than 20; the identities of most of the metabolites remain to be determined. This work illustrates that dansylation labeling and fast LC/FTICR MS can be a powerful technique for quantitative profiling of at least 672 metabolites in urine samples in 12 min. Liquid chromatography combined with mass spectrometry (LC/MS) has become an increasingly important tool for metabolome profiling.1-4 An ideal LC/MS platform would identify and quantify all metabolites present in a biological sample such as cell extracts and biofluids. Unfortunately, because of the great diversity in chemophysical properties of metabolites, it is very challenging to detect all metabolites at once. One strategy to tackle the diversity issue is to fractionate the metabolome into several groups, according to hydrophobicity, chemical structures, or other property, and then analyze them using a combination of several optimized LC/MS methods with each tailored to a group of metabolites. We are currently pursuing an analytical strategy of selectively labeling metabolites containing a certain chemical moiety, followed by LC/MS analysis for metabolite identification and quantification.5 In this work, we report a facial metabolome profiling technique for analyzing metabolites containing amines and phenolic hydroxyls or phenols. Amines and phenols are major groups of metabolites in a metabolome.6-14 Quantitative profiling of amine- and phenolcontaining metabolites in complex biological samples is important (1) Dettmer, K.; Aronov, P. A.; Hammock, B. D. Mass Spectrom. Rev. 2007, 26, 51–78. (2) Lu, W.; Bennett, B.; Rabinowitz, J. J. Chromatogr., B 2008, 871, 236–242. (3) Metz, T.; Page, J.; Baker, E.; Tang, K.; Ding, J.; Shen, Y.; Smith, R. TrAC, Trends Anal. Chem. 2008, 27, 205–214. (4) Theodoridis, G.; Gika, H.; Wilson, I. TrAC, Trends Anal. Chem. 2008, 27, 251–260. (5) Guo, K.; Ji, C.; Li, L. Anal. Chem. 2007, 79, 8631–8638. (6) Wishart, D.; Tzur, D.; Knox, C.; Eisner, R.; Guo, A.; Young, N.; Cheng, D.; Jewell, K.; Arndt, D.; Sawhney, S.; Fung, C.; Nikolai, L.; Lewis, M.; Coutouly, M.; Forsythe, I.; Tang, P.; Shrivastava, S.; Jeroncic, K.; Stothard, P.; Amegbey, G.; Block, D.; Hau, D.; Wagner, J.; Miniaci, J.; Clements, M.; Gebremedhin, M.; Guo, N.; Zhang, Y.; Duggan, G.; MacInnis, G.; Weljie, A.; Dowlatabadi, R.; Bamforth, F.; Clive, D.; Greiner, R.; Li, L.; Marrie, T.; Sykes, B.; Vogel, H.; Querengesser, L. Nucleic Acids Res. 2007, 35, D521– D526. (7) Sofic, E. Turk. J. Biochem. 2007, 32, 120–129. (8) Tsunoda, M. Anal. Bioanal. Chem. 2006, 386, 506–514. (9) Kim, Y.; Maruvada, P.; Milner, J. Future Oncol. 2008, 4, 93–102.

Analytical Chemistry, Vol. 81, No. 10, May 15, 2009

3919

for biological studies and disease biomarker discovery using metabolomics. For example, amino acids and their derivatives are common biomarkers for human physiological processes.15-21 Their identification and quantification in human fluids provides significant insight into human health. Another example is polyamines essential for eukaryotic cellular growth; rapid tumor growth is associated with polyamine biosynthesis and accumulation.22 Many studies have found significantly higher levels of polyamines and their metabolites present in the biological fluids and affected tissues of cancer patients and other patients with hyperproliferative diseases.23 Some therapeutic polyamine analogues have been shown to be potentially useful in treating cancer and other hyperproliferative disorders.22 Quantifying aminecontaining metabolites could potentially be applied to monitor tumor growth and regression in cancer studies.22,24 MS-based quantification of a large number of metabolites is not straightforward. The commonly used electrospray ionization (ESI) technique is prone to interference and ion suppression from matrix molecules or coeluting compounds during the LC/MS runs.1,25,26 For analysis of a small number of metabolites, stableisotope-labeled (SIL) analogues are often used as the internal standards to overcome the matrix and ion suppression effects.27 However, for metabolome analysis, the number of available SIL standards is very limited, and the synthesis of the SIL analogue of each metabolite will be very expensive and not practical. Instead of synthesizing an isotope analogy of the analyte of interest, the differential isotope labeling (DIL) method uses a chemical reaction to introduce an isotope tag to the analyte in one sample and another mass-difference isotope tag to the same analyte in another comparative sample (or standard), followed by mixing the two labeled samples for mass spectrometric analysis. The peak (10) Takano, K.; Ogura, M.; Yoneda, Y.; Nakamura, Y. Neuroscience 2005, 134, 1123–1131. (11) Selmaoui, B.; Aymard, N.; Lambrozo, J.; Touitou, Y. Life Sci. 2003, 73, 3073–3082. (12) Cui, Q.; Lewis, I.; Hegeman, A.; Anderson, M.; Li, J.; Schulte, C.; Westler, W.; Eghbalnia, H.; Sussman, M.; Markley, J. Nat. Biotechnol. 2008, 26, 162–164. (13) Smith, C.; O’Maille, G.; Want, E.; Qin, C.; Trauger, S.; Brandon, T.; Custodio, D.; Abagyan, R.; Siuzdak, G. Ther. Drug Monit. 2005, 27, 747–751. (14) Schauer, N.; Steinhauser, D.; Strelkov, S.; Schomburg, D.; Allison, G.; Moritz, T.; Lundgren, K.; Roessner-Tunali, U.; Forbes, M.; Willmitzer, L.; Fernie, A.; Kopka, J. FEBS Lett. 2005, 579, 1332–1337. (15) Blachier, F.; Mariotti, F.; Huneau, J.; Tome, D. Amino Acids 2007, 33, 547–562. (16) Antoniewicz, M.; Kelleher, J.; Stephanopoulos, G. Anal. Chem. 2007, 79, 7554–7559. (17) Auray-Blais, C.; Cyr, D.; Drouin, R. J. Inherited Metab. Dis. 2007, 30, 515– 521. (18) Mayboroda, O.; Neususs, C.; Pelzing, M.; Zurek, G.; Derks, R.; Meulenbelt, I.; Kloppenburg, M.; Slagboom, E.; Deelder, A. J. Chromatogr., A 2007, 1159, 149–153. (19) Shanaiah, N.; Desilva, M.; Gowda, G.; Raftery, M.; Hainline, B.; Raftery, D. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 11540–11544. (20) van Doorn, M.; Vogels, J.; Tas, A.; van Hoogdalem, E.; Burggraaf, J.; Cohen, A.; van der Greef, J. Br. J. Clin. Pharmacol. 2007, 63, 562–574. (21) Fonteh, A.; Harrington, R.; Harrington, M. Amino Acids 2007, 32, 203– 212. (22) Casero, R. A.; Marton, L. J. Nat. Rev. Drug Discovery 2007, 6, 373–390. (23) Khuhawar, M. Y.; Qureshi, G. A. J. Chromatogr., B 2001, 764, 385–407. (24) Want, E. J.; Nordstrom, A.; Morita, H.; Siuzdak, G. J. Proteome Res. 2007, 6, 459–468. (25) Cech, N. B.; Enke, C. G. Mass Spectrom. Rev. 2001, 20, 362–387. (26) Gangl, E.; Annan, M.; Spooner, N.; Vouros, P. Anal. Chem. 2001, 73, 5635– 5644. (27) Niessen, W. M. A.; Manini, P.; Andreoli, R. Mass Spectrom. Rev. 2006, 25, 881–899.

3920

Analytical Chemistry, Vol. 81, No. 10, May 15, 2009

intensity ratio of the isotope labeled analyte pair provides the basis of relative quantification of the analyte in two comparative samples or absolute quantification of the analyte in a sample if the other sample is a standard compound with a known concentration. DIL is widely used for quantitative proteome analysis.28-30 However, only a few reports are on the use of DIL for quantitative metabolome analysis. One of the early reports of using DIL for metabolite analysis was the use of the iTRAQ reagent, commonly known as the labeling reagent for peptides for quantitative proteomics, to label amino acids for quantitative analysis of these small molecules in urine and blood samples.31 Fukusaki et al. reported the use of 13 C- and 12C-methylation to introduce differential isotope tags to flavonoids for relative quantification.32 Yang et al. described an LC/MS method for amino acid analysis involving derivatization with an N-hydroxysuccinimide ester of N-alkylnicotinic acid where the alkyl chain can contain deuterium, instead of hydrogen, to provide a differential isotope tag.33 Shortreed et al. reported the use of heavy and light isotopic forms of methyl acetimidate for the relative quantification of amine-containing species.34 Guo et al. used the reductive amination reaction to label amine-containing metabolites with 13C- and 12C-formaldehyde for relative metabolome quantification.5 Ji et al. reported the use of acetaldehyde-d4 to label and quantify the monoamine neurotransmitters in rat brain microdialysates.35 Abello et al. developed isotope tagged pentafluorophenyl-activated esters of poly(ethylene glycol) to label amine-containing metabolites with multiplexing capability.36 13C4 labeled succinic anhydride and deuterated (D9) butanol have been used for labeling metabolites for relative metabolome analysis.37 While LC/MS is commonly used for detecting the differential isotope labeled metabolites, GC/MS has also been combined with chemical derivatization with isotope-coded reagents for metabolome analysis.38 It should be noted that a related method using isotope enriched media for cell culturing has been used for quantitative metabolomics.39-45 While isotope labeling can be useful for MS-based metabolome quantification, another major challenge for metabolome profiling (28) Leitner, A.; Lindner, W. Proteomics 2006, 6, 5418–5434. (29) Ong, S.; Foster, L.; Mann, M. Methods 2003, 29, 124–130. (30) Gygi, S.; Rist, B.; Gerber, S.; Turecek, F.; Gelb, M.; Aebersold, R. Nat. Biotechnol. 1999, 17, 994–999. (31) Casetta, B.; Daniels, S.; Stanick, W.; Cox, D.; Nimkar, S.; Cardenas, J.; Gamble, T. Clin. Biochem. 2006, 39, 1099–1099. (32) Fukusaki, E.; Harada, K.; Bamba, T.; Kobayashi, A. J. Biosci. Bioeng. 2005, 99, 75–77. (33) Yang, W.; Regnier, F.; Sliva, D.; Adamec, J. J. Chromatogr., B 2008, 870, 233–240. (34) Shortreed, M. R.; Lamos, S. M.; Frey, B. L.; Phillips, M. F.; Patel, M.; Belshaw, P. J.; Smith, L. M. Anal. Chem. 2006, 78, 6398–6403. (35) Ji, C. J.; Li, W. L.; Ren, X. D.; El-Kattan, A. F.; Kozak, R.; Fountain, S.; Lepsy, C. Anal. Chem. 2008, 80, 9195–9203. (36) Abello, N.; Geurink, P.; van der Toorn, M.; van Oosterhout, A.; Lugtenburg, J.; van der Marel, G.; Kerstjens, H.; Postma, D.; Overkleeft, H.; Bischoff, R. Anal. Chem. 2008, 80, 9171–9180. (37) O’Maille, G.; Go, E.; Hoang, L.; Want, E.; Smith, C.; O’Maille, P.; Nordstrom, A.; Morita, H.; Qin, C.; Uritboonthai, W.; Apon, J.; Moore, R.; Garrett, J.; Siuzdak, G. Spectroscopy 2008, 22, 327–343. (38) Huang, X.; Regnier, F. Anal. Chem. 2008, 80, 107–114. (39) Mashego, M.; Wu, L.; Van Dam, J.; Ras, C.; Vinke, J.; Van Winden, W.; Van Gulik, W.; Heijnen, J. Biotechnol. Bioeng. 2004, 85, 620–628. (40) Wu, L.; Mashego, M.; van Dam, J.; Proell, A.; Vinke, J.; Ras, C.; van Winden, W.; van Gulik, W.; Heijnen, J. Anal. Biochem. 2005, 336, 164–171. (41) Kim, J.; Harada, K.; Bamba, T.; Fukusaki, E.; Kobayashi, A. Biosci., Biotechnol., Biochem. 2005, 69, 1331–1340.

lies in the analysis of the large portion of highly polar metabolites present in a typical metabolome of cells and biofluids.1,24 Highly polar, hydrophilic compounds are poorly retained on a reversed phase (RP) LC stationary phase and will elute at or near the initial void. Sensitivity of ESI-MS detection near the void may be significantly reduced due to poor ESI desolvation as a result of the high percentage of aqueous mobile phase in the initial RP gradient runs. Severe ion suppression by coeluted polar species and salts may further decrease the ESI signal of polar analytes. To analyze the polar and ionic metabolites, LC separation based on different separation mechanisms from RPLC, such as the HILIC column, has been reported to be useful.5,46 However, the separation efficiency of HILIC columns is relatively poor for separating complex mixtures, compared to the RP column. Moreover, the use of different columns for analyzing one sample increases the overall analysis time. Thus, it is highly desirable that chemical properties such as hydrophobicility of the analyte in a sample can be altered to an extent that they all can be separated with high efficiency using RPLC compatible to ESI-MS. In addition, the detectability of these analytes by MS should be similar, averting a bias toward a certain class of analytes, thereby increasing the metabolome coverage. The alteration of the metabolite chromatographic retention properties and MS detectability may be accomplished through chemical derivatization. In this work, we report a chemical derivatization strategy based on dansylation reaction for absolute and relative quantification of amine- and phenol-containing metabolites in a complex sample. Dansylation is simple, robust, and routinely performed for many years as precolumn derivatization for quantification of amino acids, biogenic amines and phenolic hydroxyls by thin layer chromatography (TLC) and HPLC separation followed by fluorescence or UV detection.47-52 Dansylation has also been used to form derivatives of targeted analytes, followed by LC/MS analysis, for the detection of p-chlorophenol and amines,53 four phenolcontaining metabolites of a drug,54 fenfluramine and phentermine,55 and β-estradiol and estrone.56 In a recent conference report, we illustrated that endogenous human metabolites in urine could be sensitively detected by dansylation derivatization and LC/ (42) Hegeman, A.; Schulte, C.; Cui, Q.; Lewis, I.; Huttlin, E.; Eghbalnia, H.; Harms, A.; Ulrich, E.; Markley, J.; Sussman, M. Anal. Chem. 2007, 79, 6912–6921. (43) Metz, T.; Zhang, Q.; Page, J.; Shen, Y.; Callister, S.; Jacobs, J.; Smith, R. Biomarkers Med. 2007, 1, 159–185. (44) Bennett, B.; Yuan, J.; Kimball, E.; Rabinowitz, J. Nat. Protoc. 2008, 3, 1299– 1311. (45) Madalinski, G.; Godat, E.; Alves, S.; Lesage, D.; Genin, E.; Levi, P.; Labarre, J.; Tabet, J.; Ezan, E.; Junot, C. Anal. Chem. 2008, 80, 3291–3303. (46) Kind, T.; Tolstikov, V.; Fiehn, O.; Weiss, R. Anal. Biochem. 2007, 363, 185–195. (47) Seiler, N.; Deckardt, K. J. Chromatogr. 1975, 107, 227–229. (48) Seiler, N.; Knodgen, B.; Eisenbeiss, F. J. Chromatogr. 1978, 145, 29–39. (49) Loukou, Z.; Zotou, A. J. Chromatogr., A 2003, 996, 103–113. (50) Minocha, R.; Long, S. J. Chromatogr., A 2004, 1035, 63–73. (51) Zezza, F.; Kerner, J.; Pascale, M. R.; Giannini, R.; Martelli, E. A. J. Chromatogr. 1992, 593, 99–101. (52) Barrett, D. A.; Shaw, P. N.; Davis, S. S. J. Chromatogr. 1991, 566, 135– 145. (53) Quirke, J. M. E.; Adams, C. L.; Van Berkel, G. J. Anal. Chem. 1994, 66, 1302–1315. (54) Dalvie, D.; O’Donnell, J. Rapid Commun. Mass Spectrom. 1998, 12, 419– 422. (55) Kaddoumi, A.; Nakashima, M.; Wada, M.; Kuroda, N.; Nakahara, Y.; Nakashima, K. J. Liq. Chromatogr. Relat. Technol. 2001, 24, 57–67. (56) Xia, Y.; Chang, S.; Patel, S.; Bakhtiar, R.; Karanam, B.; Evans, D. Rapid Commun. Mass Spectrom. 2004, 18, 1621–1628.

MS.57 Herein we report the synthesis of an isotope coded reagent for dansylation of amines and phenols for quantitative metabolome profiling and demonstrate that differential isotope labeling via dansylation, combined with fast LC separation of labeled metabolites and Fourier-transform (FT) ion cyclotron resonance (ICR) MS, can substantially enhance the ESI sensitivity, improve chromatographic retention, and facilitate MS-based quantification and identification of potentially hundreds of amine- and phenol-containing metabolites in a complex biological sample. EXPERIMENTAL SECTION Chemicals and Reagents. All chemicals and reagents were purchased from Sigma-Aldrich Canada (Markham, ON, Canada) except those otherwise noted. The isotope compound, 13C2dimethyl sulfate, used to synthesize the isotope tagged dansylation reagent (13C-danysl chloride) was also purchased from Sigma-Aldrich. LC/MS grade water, methanol, and acetonitrile (ACN) were purchased from Thermo Fisher Scientific (Edmonton, AB, Canada). Urine samples were collected from a healthy individual and processed by adding 50% (v/v) LC/MS grade acetonitrile and then stored in a -20 or -80 °C freezer. Synthesis of Dansyl Chloride-13C2. The synthesis of 13Cdansyl chloride as a derivatizing reagent was based on a twostep procedure described by Horner and Bergmann.58,59 Figure 1A shows the synthesis scheme. In a 25 mL round-bottom flask, 0.78 g of 5-aminonaphthalene-1-sulfonic acid was added slowly in portions to 1.09 g of sodium bicarbonate in 3.5 mL of water. Then 0.77 mL of 13C2-dimethyl sulfate was added dropwise over 30 min to the stirred ice-cooled solution. The solution was warmed to 80 °C in a hot water-bath for 30 min. After cooling to room temperature, 0.46 mL of concentrated hydrochloric acid was added to the solution, and the pH was adjusted to 4. The precipitated product, 5-dimethylamino-naphthalene-1-sulfonic acid was filtered, washed with a small quantity of water, dried in the air to a constant weight, and then further dried at 120 °C in an oven. Under ice-cooling, 5-dimethylamino-naphthalene1-sulfonic acid was ground into a powder and then mixed with 0.88 g of phosphorus pentachloride. To complete the reaction, the mixture was warmed to 60 °C for 2 h under the exclusion of moisture. About 12.5 mL of ice water was then poured in. After careful neutralization with 1.75 g of sodium bicarbonate, the product was extracted with Et2O (4 × 6 mL). The organic layer was dried using sodium sulfate. The residue was purified by flush chromatography (silica gel, 20 cm × 3 cm, 60 mL of AcOEt) and further purified by a semipreparative Grace Apollo silica normal-phase HPLC column (10 mm × 150 mm, 5 µm particles). The resulting product of 13C-dansyl chloride was then dried in a SpeedVac and stored in a -80 °C freezer. The purity and confirmation of 13C-dansyl chloride was tested against the commercial 12C-dansyl chloride using LC/FTICR MS. NMR was also used to characterize the reaction products and confirm the identity and purity of the final product. Dansylation Labeling Reaction. Figure 1B shows the reaction scheme for dansylation of amine- and phenol-containing (57) Guo, K.; Ji, C. J.; Li, L. Proceedings of the 55th ASMS Conference on Mass Spectrometry and Allied Topics, Indianapolis, IN, June 3-7, 2007; poster 278. (58) Bergmann, F.; Pfleiderer, W. Helv. Chim. Acta 1994, 77, 203–215. (59) Horner, L.; Lindel, H. Phosphorus, Sulfur Silicon Relat. Elem. 1983, 15, 1–8.

Analytical Chemistry, Vol. 81, No. 10, May 15, 2009

3921

Figure 1. Reaction schemes for (A) synthesis of the isotope labeling reagent, dansyl chloride-13C2, and (B) dansylation derivatization.

compounds. The frozen urine was thawed in an ice-bath and then centrifuged for 10 min at 12 000 rpm. About 100 µL of urine supernatant or amino acids, amine, and phenolic hydroxyl standard solutions were mixed with an equal volume of sodium carbonate/sodium bicarbonate buffer (0.5 mol/L, pH 9.4) in a reaction vial. The solutions were vortexed, spun down, and mixed with 100 µL of freshly prepared 12C-dansyl chloride solution (20 mg/mL) (for light labeling) or 13C-dansyl chloride (20 mg/ mL) (for heavy labeling). The dansylation reaction was allowed to proceed for 60 min at 60 °C with shaking at 150 rpm in an Innova-4000 benchtop incubator shaker. After 60 min, mixtures were vortexed, spun down and 30 µL of methylamine (0.5 mol/ L) was added to the reaction mixture to consume the excess dansyl chloride. The solutions were again vortexed and spun 3922

Analytical Chemistry, Vol. 81, No. 10, May 15, 2009

down. After an additional 30 min of 60 °C incubation, samples were then centrifuged. The 13C-labeled mixtures were combined with their 12C-labeled counterparts for MS analysis. The reaction vials were carefully washed twice using 50 µL of LC/ MS grade MeOH, and the washing solution was added to the initial mixture to ensure dissolution and transfer of all products for MS analysis. The combined mixtures were centrifuged for 10 min at 12 000 rpm and were ready to be injected onto a RPLC column. LC/MS. The HPLC system connected to the FTICR MS or ion-trap MS was an Agilent 1100 series binary system (Agilent, Palo Alto, CA) and was modified to reduce extra system solvent volume according to an Agilent protocol (Agilent publication no. 5988-2682EN). For the fast (12 min) chromatography runs, a

reversed-phase ACQUITY BEH C18 column (2.1 mm × 50 mm, 1.7 µm particle size, 130 Å pore size) was purchased from Waters (Milford, MA). Solvent A was 0.1% (v/v) LC/MS grade formic acid in 5% (v/v) of LC/MS grade acetonitrile, and solvent B was 0.1% (v/v) LC/MS grade formic acid in LC/MS grade acetonitrile. The binary gradient elution profile was as follows: t ) 0 min, 20% B; t ) 1.5 min, 35% B; t ) 8 min, 65% B; t ) 9.3 min, 95% B; t ) 9.8 min, 95% B; t ) 10 min, 99% B. The flow rate was 170 µL/min, and the sample injection volume was 1.0 µL. For the 65 min chromatography experiment, a reversed-phase Agilent Zorbax XDB C18 column (1.0 mm × 150 mm, 3.5 µm particle size, 80 Å pore size) was purchased from Agilent. Solvent A was 0.1% (v/v) formic acid in 5% (v/v) acetonitrile, and solvent B was 0.1% (v/v) formic acid in acetonitrile. The 65 min binary gradient elution profile was as follows: t ) 0 min, 0% B; t ) 6 min, 0% B; t ) 21 min, 30% B; t ) 54 min, 90% B; t ) 65 min, 90% B. The flow rate was 50 µL/min, and the sample injection volume was 1.0 µL. The flow from RPLC was directed to the electrospray ionization (ESI) source of a Bruker 9.4 T Apex-Qe FTICR mass spectrometer (Bruker, Billerica, MA) or a Bruker Esquire ion trap mass spectrometer. All MS spectra were obtained in the positive ion mode. It was found that negative ion detection was not as sensitive as the positive ion detection for dansylated derivatives. The Esquire LC/MS system was only used for method development. All the data presented in this work were obtained using the 9.4 T FTICR mass spectrometer. RESULTS AND DISCUSSION 13 C-/12C-Dansylation Derivatization. One of the important considerations in developing labeling chemistry to tag the metabolites with differential isotopic group(s) is that the derivatization chemistry must be simple and robust. The differential dansylation labeling technique reported herein is based on well-studied derivatization chemistry.47-52 With the use of this chemistry, primary amines, secondary amines, and phenolic hydroxyls are dansylated with high yield, while tertiary amines and alkyl hydroxyls cannot be dansylated. The derivatization process is simple with mild reaction conditions and without the need of any special equipment. Because our interest is to profile as many metabolites as possible, we have examined the performance of this derivatization chemistry on a variety of metabolites. In total, 161 metabolite standards were subjected to dansylation individually and their products were analyzed by LC/ESI MS. Among them, 121 metabolites were found to be derivatized under the chosen reaction condition (see Table 1 for the list). The other 40 metabolites including amides and indole-derivatives could not be derivatized (see Table S1 in the Supporting Information). For the amides, as expected, dansyl chloride does not react with the amide nitrogen. For the indolederivatives, it appears that, if the targeted nitrogen is a part of a conjugated structure in the metabolite molecule, this nitrogen will not react with dansyl chloride. Except for this special type of amines, our study indicates that dansyl chloride is reactive with a wide range of amines and phenols. For the 121 reactive metabolites, possible impurities, side-reaction products, ESI mass spectral patterns, and RPLC retention times of the dansylated derivatives were carefully examined by LC/MS. This

compound list forms the current dansylation standard library and, as it will be discussed below, we intend to expand this library in the future for a more comprehensive metabolome analysis. For differential isotope labeling of metabolites containing amine and phenolic hydroxyl or phenol group(s), the light reagent 12Cdansyl chloride is commercially available and the heavy reagent 13 C-dansyl chloride can be synthesized according to the reaction scheme shown in Figure 1A. We have found that the dansylation labeling process itself does not introduce LC/MS background signals. After the reaction, excess dansyl chloride can be consumed by adding methylamine. The resulting dansyl methylamine does not significantly interfere with the detection of the labeled metabolites, as it can be chromatographically separated from most of the dansylated amino acids, amines, and phenols tested (see, for example, Figure 2A). We note that, for the dansylation reaction, it is crucial to keep the buffer pH at 9.4-9.5. The buffer pH should be sufficiently high so that the amine group is present in the neutral NHx (x ) 1-2) form and should be low enough to avoid reagent hydrolysis as a competitive side reaction.60 Chromatography Improvement and ESI Signal Enhancement. Because of the diverse chemical structures and chemophysical properties of metabolites, separation and detection of these compounds in a complex biological sample may not be readily accomplished using a single LC/MS method. For example, many biofluids, particularly urine, contain a large number of highly polar and poorly ESI ionizable metabolites.5,24 Efficient LC separation and ESI MS detection of these metabolites can be difficult.25,61 However, dansylation derivatization can overcome this difficulty by alternating the chromatographic retention behavior of very polar metabolites and improving the ESI responsiveness of the analytes. Figure 2A shows the base peak ion chromatograms of two separate RPLC/MS runs from the injections of a mixture of 20 free amino acids (500 pmol each) and a mixture of dansylated amino acids (5 pmol each). Two significant differences are clearly noticeable by comparing the two ion chromatograms. One is related to the chromatographic separation; the 20 dansylated amino acids are separated much better than the underivatized amino acids. This is not surprising in light of the fact that dansylation has been shown to provide better chromatographic separation in RPLC with UV or fluorescence detection than the free amino acids.47-52 A more striking difference is on the detection sensitivity. Even though the injection amount of the individual dansylated amino acids is 100-fold less than that of the free amino acids, the overall ion signal intensity of the dansylated amino acids is much greater than that of the free amino acids. The signal enhancement factors for the ESI detectable amino acids, such as Phe, Ala, Leu, and Ile, are about 1-2 orders of magnitude. For the ESIinsensitive amino acids or highly polar ones that are not retained on the RPLC column, a separate set of experiments were carried out to compare the ESI responsiveness where individual dansyl or free amino acids were continuously infused into the ESI mass spectrometer. The MS signal enhancement (60) Gros, C.; Labouesse, B. Eur. J. Biochem. 1969, 7, 463–470. (61) Cech, N. B.; Enke, C. G. Anal. Chem. 2000, 72, 2717–2723. (62) Keller, B.; Suj, J.; Young, A.; Whittal, R. Anal. Chim. Acta 2008, 627, 71– 81.

Analytical Chemistry, Vol. 81, No. 10, May 15, 2009

3923

Table 1. List of 121 Dansyl Derivative Standards and the Corresponding Metabolites (In Boldface) Identified and Quantified in a 5-Day Pooled Human Urine Sample by Using Fast LC/FTICR MS compound

retention time (min)

concn in urine (µM)

Dns-o-phospho-L-serine Dns-o-phospho-L-tyrosine Dns-adnosine monophosphate Dns-o-phosphoethanolamine Dns-glucosamine Dns-o-phospho-L-threonine Dns-6-dimethylamine purine Dns-3-methyl-histidine Dns-taurine Dns-carnosine Dns-Arg Dns-Asn Dns-hypotaurine Dns-homocarnosine Dns-guanidine Dns-Gln Dns-allantoin Dns-L-citrulline Dns-1 (or 3-)-methylhistamine Dns-adenosine Dns-methylguanidine Dns-Ser Dns-aspartic acid amide Dns-4-hydroxy-proline Dns-Glu Dns-Asp Dns-Thr Dns-epinephrine Dns-ethanolamine Dns-aminoadipic acid Dns-Gly Dns-Ala Dns-aminolevulinic acid Dns-r-amino-butyric acid Dns-p-amino-hippuric acid Dns-5-hydroxymethyluricil Dns-tryptophanamide Dns-isoguanine Dns-5-aminopentanoic acid Dns-sarcosine Dns-3-amino-isobutyrate Dns-2-aminobutyric acid Dns-Ser-Leu Dns-Pro Dns-pyridoxine Dns-Val Dns-Met Dns-Thr-Leu Dns-3-hydroxypicolinic acid Dns-salicyluric acid Dns-Trp Dns-kynurenine Dns-Gly-Leu Dns-Gly-Trp Dns-norvaline Dns-Ala-leu Dns-ethylamine Dns-4-aminobenzoic acid Dns-Ala-Trp Dns-3-aminobenzoic acid Dns-Phe

0.92 0.95 0.99 1.06 1.06 1.09 1.20 1.22 1.25 1.34 1.53 1.55 1.58 1.61 1.62 1.72 1.83 1.87 1.94 2.06 2.20 2.24 2.44 2.56 2.57 2.60 3.03 3.05 3.11 3.17 3.43 3.88 3.97 3.98 3.98 4.58 4.70 4.75 4.79 4.81 4.81 4.91 5.06 5.07 5.27 5.35 5.40 5.40 5.47 5.51 5.59 5.66 5.79 5.85 5.89 5.89 5.90 5.99 6.00 6.08 6.20